INTRODUCTION: Mixed-linkage (1→3),(1→4)-β-d-glucans (BG) reduce cholesterol level and insulin response in humans. Despite this, their role in human metabolism and a mode of action remains largely unknown.

OBJECTIVES: To investigate the effects of three structurally different BG on human fecal metabolome in a full cross-over intervention using GC-MS metabolomics.

METHODS: Over three weeks of intervention, young healthy adults received food supplemented with BG from oat, two different BG from barley or a non-fiber control in a full cross-over design. Untargeted metabolomics and short chain fatty acid analysis was performed on day three fecal samples. ANOVA-simultaneous component analysis was applied to partition the data variation according to the study design, and PLS-DA was used to select most discriminative metabolite markers.

RESULTS: Univariate and multivariate data analysis revealed a dominating effect of inter-individual variances followed by a gender effect. Weak effects of BG intake were identified including an increased level of gamma-amino-butyrate and palmitoleic acid in males and a decreased level of enterolactone in females. Barley and oat derived BG were found to influence the human fecal metabolome differently. Barley BG increased the relative level of formate in males and isobutyrate, isovalerate, 2-methylbutyrate in females. In total 15, 3 and 11 human fecal metabolites were significantly different between control vs. BG, control vs. oat BG, and barley BG vs. oat BG, respectively.

CONCLUSIONS: The study show that human fecal metabolome largely reflects individual (∼28% variation) and gender (∼15% variation) differences, whereas the treatment effect of the BG (∼8% variation) only manifests in a few key metabolites (primarily by the metabolites: d-2-aminobutyric acid, palmitoleic acid, linoleic acid and 11-eicosenoic acid).